AI Coding Glasses

Speak the intent.
Stay in the code.

CueSpec is a display-free wearable control surface for AI coding. It turns natural speech into clearer coding instructions—and keeps multiple agents within reach.

Be there from the beginning.

Get the Kickstarter launch update. Tell us what you want CueSpec to do next.

Selected ideas may earn a free pair of CueSpec glasses or a token plan. We'd love to hear yours.

How we use your information

We store your email address, feature suggestion, signup time, and email consent securely in our Cloudflare database. We use them to send CueSpec updates and review ideas. To unsubscribe or request deletion, email service@sparock.ai.

A developer wearing glasses and working with an AI coding agent in a shared office
Directional input active
Spoken intent

“Refactor the auth flow. Keep the current API stable.”

Why CueSpec

Coding moves faster when intent does.

Long prompts are often easier to explain than type. CueSpec combines directional voice capture, speech recognition, and prompt refinement so rough spoken thoughts can become focused instructions for the right coding agent.

From voice to action

Natural in. Precise out.

  1. 01 Speak naturally

    Describe the task in your own words, even when the thought is long or incomplete.

  2. 02 Clarify the prompt

    ASR and prompt refinement organize speech into a cleaner coding instruction.

  3. 03 Direct the agent

    Choose the relevant agent or window and keep the workflow moving.

Hardware interaction

Built for quiet, deliberate control.

No optical display. The interaction stays lightweight: capture your voice, select a target, and control the flow without adding another screen.

01 Directional microphone
Focuses on the wearer’s voice and reduces surrounding conversation and noise.
02 Two physical buttons
Reliable tactile actions for quick, repeatable commands.
03 Two touch sliders
Slide or tap to move between agents, windows, and interaction targets.
04 Display-free design
Keeps attention on the computer while the glasses remain a focused interaction layer.

Three working modes

One interface, across the coding day.

01

At the computer

Express longer tasks by voice, route them to the right coding agent, and reduce window switching.

Deep work

02

In shared spaces

Use low-volume, close-range speech while directional pickup limits interference from nearby voices.

Quiet input

03

Away from the desk

Check agent status and issue lightweight commands such as continue, stop, or run again.

Light control

Open platform

API & SDK
coming soon.

We’re building CueSpec as an open interaction layer for coding tools, agents, and new developer workflows.

Talk to the team

Contact us.

Questions, ideas, or a new way to build with CueSpec? Get in touch.

service@sparock.ai
Contact address ROOM 803-A02, 8TH FLOOR
HENRY CENTRE
9-15 BUTE STREET
MONG KOK
Hong Kong